Analytica Chimica Acta 478 (2003) 5966

Amperometric biosensor based on a nanoporous ZrO2 matrixBaohong Liu, Yong Cao, Dandan Chen, Jilie Kong, Jiaqi Deng

Department of Chemistry, Fudan University, Shanghai 200433, ChinaReceived 19 March 2002; received in revised form 11 November 2002; accepted 14 November 2002

Abstract

A biosensor was investigated based on the use of ZrO2 solgel matrix for enzyme immobilization in the mild condition. Thisbioceramic zirconia alcogel has been prepared by the novel alcohothermal route with a cheap inorganic salt Zr(NO3)45H2Owith several desirable features including a large surface area (about 460 m2 g1) as well as pore volume and a well-developedtextural mesoporosity, and horseradish peroxidase was selected as a model enzyme. The results of transmission electronmicroscopy (TEM) and BET measurement of the substrate showed that the as-prepared zirconia matrix has an advantageousmicroenvironment and large surface area available for high enzyme loading. The parameters affecting both the entrapment ofenzyme and the biosensor response were optimized. The resulting biosensor exhibited high sensitivity of 111A mM1 forhydrogen peroxide over a wide range of concentrations from 2.5107 to 1.5 104 mol l1, quick response of less than10 s and good stability over 3 months. 2002 Elsevier Science B.V. All rights reserved.

Keywords: Nanoporous ZrO2 matrix; Solgel; biosensor

1. Introduction

The development of new materials that can im-prove the analytical capacities of sensor devices is animportant aspect of biosensor technology. Porous in-organic solgel materials have opened up a new routefor the fabrication of chemical sensors and biosensorsdue to the attractive properties of solgel materi-als, including the simplicity of preparation, tunableporosity, low temperature encapsulation, negligibleswelling, mechanical and biodegradational stability[112]. Reviews were presented of the state-of-theart of electrochemical biosensors employing solgelmaterials [46,13,14]. Since Braun et al. [1] demon-strated the possibility of protein immobilization insolgel silica matrices. This class of ceramic mate-

Corresponding author. Fax: +86-2165641740.E-mail address: jlkong@fudan.ac.cn (J. Kong).

rials has been further applied to produce silica-basedanalysis detector. However, the recent literature seemsto show that much attention has been paid to useof conductive bioceramic metal oxide materials. Thephysical and chemical characteristics of metal oxidescan be easily manipulated, giving easy control overpolarity, pore size distribution, and electronic prop-erties. Glezer and Lev took advantage of the highconducting of vanadium pentaoxide solgel for de-signing a glucose biosensor [15]. Topoglidis et al.[16,17] reported the immobilization and bioelectro-chemistry of proteins on nanoporous TiO2 and ZnOfilms, respectively. Cosnier et al. [18] investigated aglucose biosensor within polypyrrole films electrode-posited on TiO2. Wang and Pamidi [19] reportedthe new gold-ceramic electrodes for electroanalyti-cal applications by adding the enzyme to the solgelsolution by dispersion of the metal power. Manynew amperometric biosensors based on novel solgel

0003-2670/02/$ see front matter 2002 Elsevier Science B.V. All rights reserved.PII: S0 0 0 3 -2670 (02 )01480 -0

60 B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966

composite have offered great prospects for analyticalapplications.

In addition, considering the protein adsorption onnanocrystalline films was mainly electrostatic and con-trolled by the pH, the protein charge and the solutionionic strength [16,17]. Here, we extended the study toan attractive matrix, ZrO2 by the solgel method. Itsisoelectric point is about 4.15 [20], which should bemore suitable for the adsorption of high IEP protein.

Zirconia is a technological important material thathas recently attracted considerable interest in practicalapplications [21]. It is a unique material of excel-lent chemical inertness and biocompatibility feature;moreover its surface has both oxidizing and reducingproperties, as well as acidic and basic properties [22].It has been reported that titania- and zirconia-solgelcarbon composite electrodes were used as substratesfor decrease in the overvoltage of NADH oxidationreaction [23]. Dong and Kuwanas research [24]showed that the dispersion of numerous metal oxidessuch as SiO2, ZrO2 on electrode surfaces enhancedthe oxidation rate of hydroquinone, which attributedto the adsorption of the electroactive species and theacceleration of the proton transfer step. Therefore, inelectrochemical bioanalysis, it should be studied as animportant promising candidate for catalysts and sup-port material. However, the use of highly expensivestarting material was required, which strongly limitedthe possible fabrication of a biosensor. We are notaware of any previous reports of such more conve-nient and economical methodology to prepare ZrO2matrix for protein immobilization. Here, the prepara-tion, characterization and electroanalytical utility of astable zirconia solgel-derived biosensors have beendescribed.

In this paper, we presented for the first time a novelenzyme biosensor based on a mesoporous ZrO2 ma-trix readily prepared by solgel procedure using acheap inorganic salt [Zr(NO3)45H2O] as starting ma-terial in the neutral condition. The results showed thatthis novel synthesis route affords the preparation ofthe zirconia with several desirable features includinga large surface area as well as pore volume and awell-developed porosity in the mesoporous range, pro-viding a new economical approach for protein immo-bilization. It suggested that the use of ZrO2 solgelmight be advantageous for the development of biosen-sors. The resulting biosensor exhibited high sensitivity

of 111A mM1, quick response and good stability.A comparison of the PGE/PTH/HRP zirconia solgelbased biosensor with the PGE/ZrO2 based sensor wasalso presented.

2. Experimental

2.1. Reagents

Horseradish peroxidase was purchased from Sigma(HRP, E.C.1.11.1.7, A> 250 units/mg), thionine (TH)from Fluka, Zr(NO3)45H2O and H2O2 (30% (w/v))solution was purchased from Shanghai ChemicalReagent Company. All other chemicals were of analyt-ical grade and were used without further purification.All solutions were prepared with the second-distilleddeionized water.

2.2. Apparatus

A three-electrode system was employed, with a plat-inum foil counter electrode, a saturated calomel ref-erence electrode (SCE), and the biosensor workingelectrode. Amperometric measurements were carriedout on CHI 660A (CH Instrument ElectrochemicalWorkstation, USA). A magnetic stirrer and bar witha constant-temperature control provide the convectivetransport.

Transmission electron microscopy (TEM) imagesof ZrO2 solgel matrix was taken with a HitachiH-800, using an accelerating voltage of 200 kV(Hitachi, Tokyo, Japan). Fourier transform infrared(FT-IR) spectra measured by KBr pellets containingthe solgel membranes or tyrosinase/solgel mem-branes, were obtained in the range of 2000500 cm1on an Avatar 360 spectrometer (Nicolet, USA) atroom temperature. The surface area of the sampleswere measured using the BET method by N2 ad-sorption and desorption at 77 K in a micrometricsASAP2000 system.

2.3. Preparation of zirconia solgel solution

According to our preparation [25], appropriateamount Zr(NO3)45H2O were dissolved in 50 ml ab-solute ethanol to obtain a 0.3 M solution followed byputting the solution into an autoclave of 50 ml ca-pacity. After sealed, the autoclave was maintained at

B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966 61

383 K for 60 min, and then allowed to cool to roomtemperature. A translucent alcogel was thus obtained.The ethanol gel was aged at room temperature for180 min. A stable and homogeneous ZrO2 solgel hasbeen obtained (0.4 M) and used as a stock standardsolution. This formulation could be changed whencertain experimental parameters were investigated.

2.4. Construction of the H2O2 biosensor

A 4 mm pyrolytic graphite electrode (PGE) wasused as the base electrode for the solgel-modifiedH2O2 biosensor. The PGE was polished by 0.3, 0.1,0.05m alumina particles, rinsed thoroughly withdeionied acetone and double-distilled water in anultrasonic bath, and dried in air before use.

Firstly, the electropolymerization of thionine wascarried out in two steps following our previous re-search. Holding the PGE under a constant potentialof 1.5 V for 5 min and voltammetric cycles between0.45 V and 0.05 V at 100 mV s1 in pH 6.5 phos-phate buffer solution containing 0.1 mM thionine.All solutions tested were thoroughly deoxygenatedby bubbling pure nitrogen, and a continuous flowof nitrogen was maintained over the solution duringexperiments unless stated otherwise.

Then the immobilization of HRP in the ZrO2solgel matrix was accomplished by addition of 10lof HRP (10 mg ml1) completely with 10l ZrO2solgel. With a micropipet, 10l aliquots of such acolloid were deposited on the PGE surface and dryingunder 4 C in air. Then the biosensor was prepared(PGE/PTH/HRP-solgel). The sensor was kept inair at 4 C between measurements. As the compari-son, PGE/ZrO2 electrode without enzyme was alsoprepared.

2.5. Measurement procedure

Cyclic voltammetric and amperometric measure-ments were performed with CHI 660 electrochemicalworkstation (CHI, Cordiva, USA). The three-electrodesystem was used for detection. All the experimentalsolutions were thoroughly deoxygenated by bubblingnitrogen through the solution all over the experiment.

In the constant potential experiments, successiveadditions of stock H2O2 solution to the phosphatebuffer were made and the currenttime data were

recorded as a function of time following the addition ofH2O2.

3. Results and discussion

3.1. Interaction between ZrO2 solgel and HRP

ZrO2 solgel has a porous structure of matrixfor embedding the enzymes. To investigate the mi-crostructures of the as-synthesized ZrO2 matrix andthe immobilization of enzyme in the matrix, transmis-sion electron microscopy images were experimented.From Fig. 1, it can be seen that the matrix was highlyporous in nature, which consisted of clusters ofcross-linked particles about 5 nm. The TEM picture ofZrO2 matrix also displayed a three-dimensional uni-form porous structure, which provided a significantincrease of effective electrode surface for enzymeloading and good preparation reproducibility. Thesurface area and the pore volume of the matrix, mea-sured using the BET method by N2 adsorption anddesorption, were about 460 m2 g1 and 2.9 cm3 g1,respectively.

The interactions between ZrO2 matrix and HRPwere studied by infrared spectra. The matrix has adsor-ption bands at 781 and 1054 cm1, characteristics offrame asymmetric and symmetric flexible vibrations,

Fig. 1. The transmission electron microscopy image of the as-syn-thesized ZrO2 matrix and the immobilization of enzyme in thematrix.

62 B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966

and HRP has specific adsorption bands at 729 and1651 cm1. When both of them mixed, the specificadsorption bands of HRP still exist, while some ofbands of matrix nearly disappeared and the band at1640 cm1 shifted in the IR spectrum of ZrO2-HRP.These results sowed that there might be intermolecularinteraction between enzyme and some specific sites ofmatrix, and apparently the ZrO2 solgel was a goodimmobilization matrix for enzyme loading.

3.2. Fabrication of the biosensor

As in our previous paper [26] about the electropoly-merization of thionine on glassy carbon electrode,the pyrolytic graphite electrode was preanodized at aconstant potential of 1.5 V for 5 min, then was elec-tropolymerized in pH 6.5 phosphate buffer solutioncontaining 0.1 mM thionine from 0.05 V to 0.45 V.The cyclic voltammogram (Fig. 2) indicated thegrowth process of the PTH films with consecutivecycles. When the PTH/PGE electrode was washedthoroughly and then removed to blank phosphatebuffer solution without thionine, the current existed,which confirmed that the polymer film has formed onthe PGE surface.

The cyclic voltammogrames of the thionine modi-fied electrode at various scan rates were studied. Thepeak currents were linearly proportional to the scanrates, which indicated a thin film adsorption redox be-havior. In addition, the experiment results showed that

Fig. 2. The cyclic voltammograms of PTH progressive course ona preanodized the pyrolytic graphite electrode at 1.5 V for 5 minin pH 7.0 buffer containing 0.1 mM TH.

electrochemical behavior of the modified electrodewas influenced by pH, and the E0 value decreased by55 mV/pH between pH 6 and 8.5 for the polymerizedthionine, which was closed to the Nernstain value of59 mV for a two-electron, two-proton process.

Additionally, the film thickness, pH and the HRP/ZrO2 concentration are the important factors to in-fluence the properties of the matrix structure and theperformance of the biosensor. The thickness of thesolgel film and the response of the immobilized en-zyme were affected by varying the ratio of ZrO2 stockstandard solgel to water and ZrO2/HRP. The resultsin Fig. 3 showed that the sensitivity and dynamic re-sponse range of the biosensor increased when the ra-tio of ZrO2/H2O was smaller than 1:2. If the ratio waslower than 1:8, the response of the biosensor grad-ually decreased. Thus, the optimum ratio was fixedat 1:8 for the immobilizing HRP, and the surface ofthe solgel-modified enzyme electrode appeared to besmooth and firm under such condition. Then the effectof the HRP (10 mg ml1)/ZrO2 ratio on the biosensorresponse was further studied and the results showedthat the optimum ratio was fixed at 1:1 for the prepa-ration of the biosensor.

The pH of the solgel fabrication process mightinfluence the performance of the biosensor. The zeropoint charge values of the supports are of importancein the immobilization of enzymes since at pH val-ues above the ZPC the surface of the matrix wouldbe negatively charged, and conversely, at pH values

Fig. 3. Optimization of the concentration of ZrO2 solgel (0.4 MZrO2:HO2) on the response of the biosensor.

B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966 63

below the ZPC the surface would be positivelycharged. It has been reported that the isoelectronicpoint of Horseradish peroxidase is 8.9, while ZrO2has an IEP measurement of 4.18 [20]. Our exper-iments showed that the optimum response of thebiosensor was obtained at about pH 67 for immobi-lizing HRP, which was in good agreement with theoptimum activity of HRP.

3.3. Amperometric response of the biosensor

Fig. 4 showed the cyclic voltammogrames of theH2O2 biosensor in a deoxygenated phosphate buffersolution without H2O2 (a) and with H2O2 (b). Inthe absence of H2O2, the biosensor gave no re-sponse and only the typical electrochemical behaviorof the electropolymerized thionine was observed.Addition of H2O2 in buffer solution, the reductioncurrent increased with no increase in the oxida-tion current, which indicated that a catalytic reac-tion occurred on the biosensor. This result showedthat thionine could effectively shuttle electrons be-tween the redox center of HRP and the electrode.The process of electron transfer between H2O2and the biosensor was illustrated in the followingparagraph.

Fig. 4. Cyclic voltammograms of the biosensor in: (a) the absence;and (b) presence of H2O2 at a scan rate of 100 mV/s.

At first, HRP reduced H2O2 diffusing from solutioninto the membrane and oxidizes HRP to HRP-I

H2O2 + HRP HRP-I + H2OThe reduction of HRP-I to HRP includes two sep-

arate steps

HRP-I + PTHH HRP-II + PTH

HRP-II + PTH HRP + PTH+

where HRP-I and HRP-II are two unstable intermedi-ates. PTHH and PTH+ represent reduced and oxidizedforms of poly(phenazine methosulfate). Subsequentlyoxidized PTH+ can be recycled to be reduced to PTHHat the electrode, bring about a reduction current

PTH+ + H+ + 2e PTHHThe pH dependence of the electrocatalytic current

for the biosensor decided in two aspects: one was thepH effect of the activity of HRP; the other was thatof the peak potentials of thionine. Considering thesetwo aspects, the effect of pH on the biosensor wastested and an optimum pH 6.8 was selected in thiswork (Fig. 5), which was in agreement with that sol-uble HRP reported before [27]. Therefore the ZrO2matrix did not change the optimal pH value for thebioelectrocatalytic reaction of the immobilized HRPto H2O2.

The potential dependence of the amperometricsignal and background current of the biosensor was

Fig. 5. Effect of pH on the H2O2 biosensor at 25 C.

64 B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966

Fig. 6. A typical currenttime response curve for successive addition of 25l, 1 mmol l1 H2O2 to 10 ml buffer solution kept under stirringfor an enzyme electrode fabricated under optimized condition. Operation potential 0.2 V vs. SCE; insert curve is a calibration of thebiosensor.

tested. The results showed that the highest ration ofsignal-to-background was achieved at 0.2 V. Thelow operational potential minimized interferencesfrom other coexisting electroactive species.

Fig. 6 displayed a typical amperometric response ofthe biosensor after the addition of successive aliquotsof H2O2. A defined reduction current proportional tothe H2O2 concentration was observed. The time wasless than 10 s to reach 95% of the steady-state value,which was attributed to the fast diffusion of the sub-strate in the porous film. The biosensor exhibited alinear calibration range from 2.5 107 to 1.5 104 mol l1 with a sensitivity of 111A mM1 anda regression equation: I (A) = 0.081+0.111C(M)(in Fig. 7) and log I (A) = 2.24+0.87 log C(M) (inFig. 6 insert curve), respectively. The detection limitwas 1.0 107 mol l1 at a signal-to-noise ratio of3. Such a high sensitivity may be attributed to the fa-vorable microenvironment of the ZrO2 matrix, which

has a general affinity for the protein loading and couldstabilize its biological activity to a large extent.

In order to demonstrate the catalytic function ofZrO2 to H2O2, the amperometric characterization ofthe ZrO2-HRP/PGE and ZrO2/PGE were compared.From Fig. 7 and insert curve, the sensitivity of the for-mer was about 10 times greater than that of the latter.The ZrO2/PGE electrode exhibited a linear range from2.0104 to 3.5102 mol l1 with a detection limitof 1.0 106 mol l1.

The developed biosensor based on the ZrO2 ma-trix displayed good reproducibility. The relative stan-dard deviation determined by 20 successive assays ofa 5.0 105 mol l1 H2O2 sample was 2.7%. As toa series of six biosensors, which made independentlyunder the same conditions, a relative standard devia-tion of 7.5% was obtained for the individual currentresponses to 5.0 105 mol l1. In addition, the ac-curacy of the sensor was evaluated by determining

B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966 65

Fig. 7. Calibration curves of the ZrO2-HRP/PGE and ZrO2/PGE electrode (insert curve).

the recoveries of hydrogen peroxide after standard ad-dition of concentrations ranging from 8.5 106 to7.5104 M in sterilized milk samples. The biosensorshowed satisfactory results with an average recoveryof 100.3%.

The operational stability was investigated by con-secutive measurements of a 7.5 105 mol l1 H2O2sample. After 160 measurements in 5 h, the biosensorretained about 82% of its initial activity, and no swellof the HRP/ZrO2 solgel membrane was observedin the solution during this continuous measurements.Additionally, the long-term stability of the ZrO2solgel-derived biosensor was studied. The responseof the biosensor was tested day after day duringthe first 2 weeks, afterwards the experiments wereperiodically carried out over a period of 3 monthsand dry storage at 4 C. The results showed that theresponse decreased about 4.5% of its initial sensitiv-ity after 2 weeks and remained 75% after 3-monthusage, while the dynamic linear range of the biosen-sor decreased gradually with the storage time. Thislong-term stability was attributed to the interactionbetween HRP and the matrix. The amine and car-boxyl groups of the enzyme might interact with some

specific sites of the ZrO2 substrate, which preventthe enzyme from leaking out of the ZrO2 solgelmembranes.

4. Conclusion

The biosensor based on immobilization of HRPin a ZrO2 solgel matrix on an electropolymerizedthionine modified electrode offered advantage char-acteristics, including high sensitivity and long-termstability. A novel and easy alcohothermal process hasbeen developed to prepare the stable zirconia solgelusing a cheap inorganic salt as the starting materialwith several desirable characteristics including a largesurface area and a well-developed porosity in themesoporous range, providing a favorable microenvi-ronment for enzyme binding. The results showed thatthe zirconia solgel was an attractive matrix for en-zyme immobilization and electroanalytical measure-ments. No loss of enzyme was observed in the solgelmembrane and enzyme activity was maintained. Itcould be extended to other proteins to fabricate thebiosensors.

66 B. Liu et al. / Analytica Chimica Acta 478 (2003) 5966

Acknowledgements

This work was supported by the National NatureScience Foundation of China, the ElectroanalyticalChemistry Key State Laboratory of Changchun Insti-tute of Applied Chemistry, Shanghai Qi-ming-xingproject (01QA14007).

References

[1] S. Braun, S. Rappoport, R. Zusman, D. Avnir, M. Ottolenghi,Mater. Lett. 10 (1990) 1.

[2] D. Avnir, S. Braun, M. Ottolenghi, in T. Bein (Ed.), Procee-dings of the ACS Symposium Series, No. 499 on SyntheticControl in Thin Films and Solids, American Chemical Society,Washington, DC, 1992, pp. 384404.

[3] L.M. Ellerby, C.R. Nishida, F. Nishida, S.A. Yamannaka, B.Dunn, J.S. Valentine, J.I. Zink, Science 255 (1992) 1113.

[4] O. Lev, M. Tsionsky, L. Rabinovich, V. Glezer, S. Sampath,I. Pankratov, J. Gun, Anal. Chem. 67 (1995) 23A.

[5] O. Lev, Analysis 20 (1992) 543;S. Sampath, O. Lev, Anal. Chem. 68 (1996) 2015.

[6] B.C. Dave, B. Dunn, J. Valentine, J.I. Zink, Anal. Chem. 66(1994) 1120A.

[7] D. Avnir, S. Braun, O. Lev, M. Ottolenghi, Chem. Mater. 6(1994) 1605.

[8] R.A. Dunbar, J.D. Jordan, F.V. Bright, Anal. Chem. 68 (1996)604.

[9] J. Wang, P.V.A. Pamidi, D.S. Park, Anal. Chem. 68 (1996)2705.

[10] U. Narang, P.N. Prasad, F. Bright, K. Ramanathan, N. Kumar,B. Malhorta, M. Kamalasana, S. Chandra, Anal. Chem. 66(1994) 3139.

[11] P. Audebert, C. Demaille, C. Sanchez, Chem. Mater. 5 (1993)911.

[12] Z. Liu, B. Liu, J. Kong, J. Deng, Anal. Chem. 72 (2000)4707.

[13] J. Wang, Anal. Chim. Acta 399 (1999) 21.[14] J. Lin, C.W. Brown, Trends Anal. Chem. 16 (1997) 200.[15] V. Glezer, O. Lev, J. Am. Chem. Soc. 115 (1993) 2533.[16] E. Topoglidis, A.E.G. Cass, G. Gilardi, S. Sadeghi, N.

Beaumont, J.F. Durrant, Anal. Chem. 70 (1998) 5111.[17] E. Topoglidis, A.E.G. Cass, B. ORegan, J.R. Durrant, J.

Electroanal. Chem. 517 (2001) 20.[18] S. Cosnier, A. Senillou, M. Gratzel, P. Comte, N.

Vlachopoulos, N. Renault, C. Martelet, J. Electroanal. Chem.469 (1999) 176.

[19] J. Wang, P.V.A. Pamidi, Anal. Chem. 69 (1997) 4490.[20] T. Klimova, M.L. Rojas, P. Castillo, R. Cuevas, J. Ramirez,

Microporous Mesoporous Mater. 20 (1998) 293.[21] D.T. On, D. Desplantier-Giscard, C. Danumah, S. Kaliaguine,

Catal. Today 20 (1994) 199.[22] M.A. Aramendia, V. Borau, C. Jimenez, J.M. Marinas, A.

Marinas, A. Porras, F.J. Urbano, J. Catal. 183 (1999)240.

[23] J. Wang, P.V.A. Pamidi, M. Jiang, Anal. Chim. Acta. 360(1998) 171.

[24] S. Dong, T. Kuwana, J. Electrochem. Soc. 131 (1984) 813.[25] J. Hu, Y. Cao, J. Deng, Chem. Lett. 5 (2001) 398.[26] R. Yang, C. Ruan, W. Dai, J. Deng, J. Kong, J. Electroanal.

Chem. 44 (1999) 1585.[27] A.C. Maehly, Methods Enzymol. 11 (1955) 807.

Amperometric biosensor based on a nanoporous ZrO2 matrixIntroductionExperimentalReagentsApparatusPreparation of zirconia sol-gel solutionConstruction of the H2O2 biosensorMeasurement procedure

Results and discussionInteraction between ZrO2 sol-gel and HRPFabrication of the biosensorAmperometric response of the biosensor

ConclusionAcknowledgementsReferences